Prevalence and associated factors for alcohol use disorder among tuberculosis patients: a systematic review and meta-analysis study.

Background: Alcohol use disorders (AUD) in tuberculosis patients are complicated with poor compliance to anti-tuberculosis treatment and poor tuberculosis treatment outcomes. However, aggregate data concerning this problem is not available. Therefore, this review aimed to fill the above gap by generating an average prevalence of AUD in tuberculosis patients.
Method: Our electronic search for original articles was conducted in the databases of Scopus, PubMed, and EMBASE, African Index Medicus, and psych-info. Besides, the reference list of selected articles was looked at manually to have further eligible articles for the prevalence and associated factors of AUD in tuberculosis patients. The random-effects model was employed during the analysis. MS-Excel was used to extract data and stata-11 to determine the average prevalence of AUD among tuberculosis patients. A sub-group analysis and sensitivity analysis were also run. A visual inspection of the funnel plots and an Eggers publication bias plot test were checked for the presence of publication bias.
Result: A search of the electronic and manual system resulted in 1970 articles. After removing duplicates and unoriginal articles, only 28 articles that studied 30,854 tuberculosis patients met the inclusion criteria. The average estimated prevalence of AUD in tuberculosis patients was 30% (95% CI: 24.00, 35.00). This was with a slight heterogeneity (I ²  = 57%, p-value < 0.001). The prevalence of AUD in tuberculosis patients was higher in Asia and Europe; 37% than the prevalence in the US and Africa; 24%. Besides, the average prevalence of AUD was 39, 30, 30, and 20% in studies with case-control, cohort, cross-sectional and experimental in design respectively. Also, the prevalence of AUD was higher in studies with the assessment tool not reported (36%) than studies assessed with AUDIT. AUD was also relatively higher in studies with a mean age of ≥40 years (42%) than studies with a mean age < 40 years (24%) and mean age not reported (27%). Based on a qualitative review; the male gender, older age, being single, unemployment, low level of education and income from socio-demographic variables, retreatment and treatment failure patients, stigma, and medication non-adherence from clinical variables were among the associated factors for AUD.
Conclusion: This review obtained a high average prevalence of AUD in tuberculosis patients and this varies across continents, design of studies, mean age of the participants, and assessment tool used. This implied the need for early screening and management of AUD in tuberculosis patients.